HAL Id: cea-02510729https://hal-cea.archives-ouvertes.fr/cea-02510729
Submitted on 18 Mar 2020
HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.
Molecular dynamics simulation of ballistic effects insimplified nuclear waste glasses
Amreen Jan, Jean-Marc Delaye, S. Gin, Sebastien Kerisit
To cite this version:Amreen Jan, Jean-Marc Delaye, S. Gin, Sebastien Kerisit. Molecular dynamics simulation of ballisticeffects in simplified nuclear waste glasses. Journal of Non-Crystalline Solids, Elsevier, 2019, 505,pp.188-201. �10.1016/j.jnoncrysol.2018.11.021�. �cea-02510729�
Molecular dynamics simulation of ballistic effects in simplified
nuclear waste glasses
Amreen Jan1, 2
, Jean-Marc Delaye1, Stéphane Gin
1, Sebastien Kerisit
2
1CEA, DEN, Service d’études de vitrification et procédés hautes températures, 30207
Bagnols-sur-Cèze, France 2 Physical Sciences Division, Pacific Northwest National Laboratory, Richland, Washington
99352, United States
Abstract
Ballistic effects in simple sodium borosilicate (
) and sodium alumino-borosilicate
glasses (
) were investigated using molecular dynamics simulations.
Specifically, the glasses were irradiated with heavy projectiles that caused atomic
displacements by elastic collisions (displacement cascades) and progressively damaged the
bulk glass. The accumulated pressure and internal energy inside the glass were found to
saturate with deposited energy. Furthermore, structural analysis of the irradiated glasses
revealed several important ballistic effects including a decrease in glass density,
depolymerization of the borosilicate network, and increase in chemical mixing, short range
and intermediate disorder. The magnitude of radiation damage was found to depend on the
glass composition and, in general, alumino-borosilicate glasses were found to be slightly less
damaged, after irradiation, as compared to borosilicate glasses.
Keywords
Molecular dynamics, borosilicate glasses, ballistic effects
1. Introduction
Immobilization of High Level Waste (HLW) in borosilicate matrices and further disposal into
geological repository has been regarded as one of the best ways for long-term isolation of
HLW from the biosphere [1,2]. To confine radionuclides (RN), a glass must present a high
physical and chemical durability. Under repository conditions, RN are expected to stay
confined until the glass matrix has been reached by groundwater and begins to corrode [3].
Consequently, glass corrosion has been studied extensively, focusing primarily on the effects
of environmental conditions (pH, temperature, glass-surface-to-solution-volume ratio) and
glass compositions, and different regimes of alteration have been identified (initial alteration
rate, residual rate, and resumption rate [4,5]) . Today, it is widely accepted that the alteration
layer that forms between the pristine glass and the solution is responsible for the alteration
rate drop [6–8]. But the glass corrosion behavior, and thus glass durability, may be affected
by internal emission of radiation particles (α, β, and γ).
Various studies have been conducted to determine the effects of each type of radiation on the
structure of unaltered glasses using doping techniques (Cm, Pu), external irradiation beams,
and atomistic modeling (for ballistic collisions only) [9–13]. Due to their long half-lives, α
radionuclides are expected to dominate the long-term contribution to the deposited energy
[14] and, indeed, a study of nuclear waste glasses subjected to different irradiation fluxes
showed that recoil nuclei produced during α-decay were the main source of irradiation
damage [15]. The nuclear collisions (or ballistic collisions) lead to swelling, a decrease of
hardness, an increase in fracture toughness, and an increase in structural disorder and
depolymerization [15].
The extent by which radiation can affect glass corrosion, however, is still controversial.
Earlier studies conducted on the impact of radiation (α, β, and γ) reported no significant
impact on glass alteration kinetics (both initial and residual rates) [16–21]. In contrast, recent
experiments conducted by Mougnaud et al. [22] in residual rate regime, using multiple energy
gold ion irradiation with dominant nuclear dose(~200MGy), showed a significant increase of
the alteration rate and thickness of the alteration layer in pre-irradiated glasses compared to
non-irradiated glasses. In fact, changes in the residual rate and the alteration layer thickness
appeared to attain saturation around the same deposited dose, and thus showed a behavior
akin to that described above for structural properties. Therefore, it remains unclear to what
extent radiation damage to the pristine glass structure can impact corrosion. In particular, the
correlation between the alteration layer structure and the initial glass structure remains an
open question.
The present work is the first part of an ongoing study aimed at addressing this question using
atomistic modeling techniques, namely, classical molecular dynamics (MD) and Monte Carlo
(MC) methods [23–25]. Here, series of pristine and irradiated sodium borosilicate (SBN) and
sodium alumino-borosilicate (SBNA) glasses were simulated using MD and a systematic
analysis of radiation effects was performed to correlate structural changes induced by the
ballistic effects to glass chemical composition. In the second part of this study, which will be
the subject of a future publication, the structural changes will be implemented in a Monte
Carlo code to investigate their effects on glass corrosion.
MD simulations have proved to be a powerful tool for understanding ballistic damage induced
by recoil nuclei in glasses, particularly the changes in short- and medium-range order[26,27]
[27,28]. Density evolution with deposited dose using MD followed similar trends as measured
for curium-doped and externally-irradiated glasses. In fact, stabilization of glass swelling
around same dose (1021
keV.cm-3
) bolstered the idea that network modifications are a
consequence of damage induced by recoil nuclei [11]. Furthermore, correlation of
microstructural parameters (increase in disorder and depolymerization) and experimental
results has led to the “ballistic disordering fast quenching model”. In this model, displacement
cascades (DC) generated by the recoil nucleus create significant structural disorder and also
raise the local temperature and, since the bulk of the glass is at ambient temperature, it
quickly quenches from high fictive temperature. This occurs at very short time scales (10-12
s)
and the glass is thus not able to reach an equilibrium state and freezes at higher enthalpy [28–
31].
Reliable interatomic potentials are critical to an accurate description of the structure and
properties of pristine and irradiated glasses. In particular, a limited number of potentials are
available for SBN and SBNA glasses due to the difficulties in simulating accurately the non-
linear dependence of boron coordination on composition[32,33]. The correct reproduction of
boron coordination is an important criterion to consider while developing new potentials. A
few years ago, Kieu et al.[34] developed a set of composition-dependent potentials for SBN
glasses by fitting parameters based on Dell and Bray model [32,35]. Even though these
potentials have been criticized because of the charge-composition dependencies that could be
considered as unphysical, good agreement with experiment was obtained for the structural and
mechanical properties, within the domain of validity of these potentials. However, depending
on the way the glass is prepared, these potentials can overestimate the density[36]. Several
other potentials have been derived to simulate borosilicate glasses [37–42] but their use has
been limited due to either their complexity or their failure to simulate some compositions.
Three Buckingham-based potentials, proposed by Jolley et al. [43], Deng et al. [33] or Stoch
et al.[44] were tested in this work. The results obtained with these different potentials have
been compared to explain the potential selected for the DC simulations.
This work is organized as follows: Section 2 presents the methodology used for simulating
pristine and irradiated glasses and further methods used for the analysis of results. Section 3
presents the results, first on the potential selection and then results from DC. Finally, SBN
and SBNA glasses are compared and Section 4 summarizes our findings and discusses their
significance.
2. Materials and methods
2.1 Interatomic potentials
The simulations presented in this study were performed using the molecular dynamics code
DL_POLY [45]. The interactions between the atoms (Si, B, O, Na and Al) have been
calculated using potentials of the form shown below, comprising a long-range Coulomb term
and a short-range Buckingham term:
(1)
Where is the distance between two atoms i and j, are the effective charges of the
atoms, and , , are parameters describing repulsion and dispersion interactions between
two atoms. Coulombic interactions were treated by the Ewald summation method[46].
Parametrizations of the Buckingham potentials proposed by Jolley et al. [43], Stoch et al.[44]
and Deng et al. [33] were selected from literature for comparison (the Table enumerating the
pair potential parameters can be found in supporting information). Further, depending upon
the accuracy of the simulation with respect to experiment, one potential was selected for the
glass preparation and the radiation effect simulations.
2.2 Glass preparation scheme and presentation of the chemical compositions
Pristine sodium borosilicate (SBN) and sodium alumino-borosilicate (SBNA) glasses were
simulated following the scheme shown in Fig.1. A cubic cell containing 45000 randomly
placed Si, B, O, Na and Al atoms with an initial density determined by the method detailed in
Section 2.2 was first equilibrated at 4000 K for 100000 time steps (1 time step = 1 fs) in NVT
ensemble. The structure was then quenched to room temperature (300 K) by decreasing the
temperature 100 K per step in order to apply a quench rate of 5 K/ps. After quenching, the
resulting glass was relaxed at room temperature and pressure in the NPT ensemble for 20000
time steps. Final relaxation of 5000 steps in NVE ensemble was applied. Periodic conditions
were systematically applied throughout the simulation.
Figure 1. Glass preparation scheme by MD simulation
The following ratios are used throughout this work to define the glass compositions:
Table 1 shows the ten glass compositions that were investigated. Four SBN compositions with
R = 1 and varying K and four SBNA compositions with Rˊ = 1 and varying Kˊ were prepared.
These compositions were selected specifically because aqueous alteration of such
compositions has been studied by Monte Carlo simulations. Two more glasses, SBN-14 (also
known as CJ1) and SBNA-CJ2 were also prepared. These glasses are closest (in terms of R
and Rˊ ratios) to R7T7 glass – the French nuclear waste glass – and have been widely studied
experimentally for irradiation effects and aqueous alteration.
Table 1. Glass compositions investigated here, except for 5th
and 8th
compositions, all other
possess and
S.no Glass Name %SiO2 %Na2O %B2O3 %Al2O3 K or Kˊ
1. SBN-955 90 5 5 - 18
2. SBN-15 70 15 15 - 4.67
3. SBN-622 60 20 20 - 3
4. SBN-433 40 30 30 - 1.33
5. SBN-14/CJ1 67.73 14.23 18.04 - 3.75
6. SBNA-2 66 17 15 2 3.88
7. SBNA-5 60 20 15 5 3
8. SBNA-CJ2 61.2 13.3 18.9 6.6 2.4
9. SBNA-10 50 25 15 10 2
10. SBNA-12 46 27 15 12 1.7
2.3 Choice of the initial glass density for glass preparation
A recent study by Kilymis et al. [10] revealed the importance of the choice of the initial
density on the final glass density. Consequently, depending on the glass density, swelling or
contraction can be observed after series of displacement cascades. It was observed that when
the pristine glass is prepared with an initial density corresponding to the minimum of the
potential energy (PE), the radiation effects are in best agreement with experimental results. A
similar study by Jolley et al.[36] using current potentials, revealed that the simulated glasses
with experimental density actually may not correspond to the glass with minimum PE.
Hence, in order to make sure the simulated pristine glasses correspond to a minimum of the
PE, a specific algorithm was applied. In this algorithm, a glass composition (4000 atoms) was
prepared with different densities (+/-15% the experimental one) using the preparation scheme
shown in Fig. 1 but without the relaxation in the NPT ensemble (i.e. the volume was kept
constant during the whole glass preparation). Afterwards, the potential energy and pressure of
the equilibrated glass were plotted versus density and the curves were fitted to determine two
equilibrium densities corresponding to the minimum PE and zero pressure, respectively (see
Fig. 1). The average of the two minima, which were always very close, was taken as the
effective equilibrium density. Finally, the initial density used for the pristine glass preparation
(presented in Section 2.1) was taken to be 5% larger than the effective equilibrium density to
account for the slight swelling that occurs during the NPT stage at the end of the preparation.
Figure 2.Pressure (black line) and total PE (blue) for SBN- when different
initial densities are imposed. Arrows represent the equilibrium densities
corresponding to minimum PE and zero pressure. Average of these two
was taken as the effective equilibrium density
2.4 Simulation of series of displacement cascades
To simulate the radiation effects, series of displacement cascades were initiated for cubic
simulation cells of the SBN and SBNA glasses containing 45000 atoms and with an average
edge length close to 78 Å. An edge length of 78 Å was found to be sufficient with respect to
path length of projectile (confirmed by software SRIM[47]). Before launching a series of DC
in a glass, 8 Si atoms were randomly assigned the mass of uranium (these pseudo U
projectiles were otherwise treated exactly like Si atoms), and then the glass was equilibrated
in NPT and NVE ensembles. The series of DC was then started by choosing U projectile and
placing it at one of the corners and accordingly, the simulation cell was translated by the
vector formed by the final and initial positions of the projectile. The projectile was then
launched towards the center of the cell with an initial kinetic energy of 4 keV. After the first
projectile was completely stopped, another projectile was selected and placed at another
corner, and the procedure was repeated cycling through the 8 corners until 120 cascades are
simulated. This approach ensures that the whole simulation box is progressively irradiated by
the projectiles. After the last cascade, equilibrations were performed in the NPT and NVE
ensembles to relieve any built-up pressure and to determine the new equilibrium volume.
To represent more precisely the atomic interactions that can occur at the short range between
projectile and the glass atoms during a displacement cascade, Zeigler-Biersack-Littmark
(ZBL) potentials[48] have been used. Furthermore, to ensure the continuity of energy, forces,
and force derivatives between ZBL and Buckingham potentials, complementary potential with
a polynomial form was introduced.
2.5 Analysis
2.5.1 Non-bridging oxygens per species
To study the change in the polymerization level, the numbers of non-bridging oxygens (NBO)
per species before and after DC were compared. The number of NBO per species was
calculated using the formula given below:
(1)
Where Qn represents fourfold-coordinated Si/B/Al atoms with n bridging oxygens (BO) and 4
− n NBOs. Similarly for the case of trigonal boron,
(3)
Where Tn represents threefold-coordinated B atoms with n BOs and 3− n NBO.
2.5.2 Number of triplets and comparison to chemical random mixing
To quantify the degree of chemical ordering in each glass, the number of triplets obtained
assuming random mixing were compared to the number of triplets in the simulated glass.
Number of triplets on random mixing are calculated as follows [49]. If corresponds to the
number of two coordinated O, the number of F1-2O-F2 triplets (F1 and F2 are two network
former atoms) that would be obtained in the case of a chemical random mixing is given by:
F1-2O-F2 =
(2)
is the probability of forming one F1-2O-F2 triplet in a structure containing bonds of
type F1-2O and bonds of type F2-
2O. Two cases must be distinguished.
Case 1: F1 and F2 are different former types
= 2*
(5)
Where,
Case 2: F1 and F2 are the same former type
=
(3)
3. Results
3.1 Force field selection
Structural properties of the SBN-14 glass prepared with each potential were compared with
experimental results (Table 2). The Jolley and Stoch potentials underestimated the glass
density by 8% and 2%, respectively, whereas the Kieu potential overestimated the density by
11%. The average boron coordination was poorly estimated by the Jolley and Stoch potential,
the Kieu potential gave an estimate within 1.5% of the experimental value. Other structural
properties such as the average bond angle and bond length were also best estimated by the
Kieu potential.
Hence, all the glasses in Table 1 were prepared with the Kieu potential for the ternary glasses
and the Deng potential for the quaternary glasses (the Deng and Kieu potentials are strictly
identical for the ternary glasses and the Deng potential allows for Al2O3 in the composition to
be simulated). SBN-14, in particular, was prepared with both the Kieu and Jolley potentials to
study how radiation effects may be impacted by the nature of the potential.
Table 2. SBN-14 (67.73%SiO2.14.23%Na2O.18.04%B2O3) prepared with three different
potentials. This table compares results from three potentials with experimental results. (#CB
represents mean boron coordination)
Parameter Simulation Results Experimental
Results[35] Jolley
Potential
Stoch
Potential
Kieu
Potential
Equilibrium
Density(g/cm3)
2.25 2.50 2.73 2.45
#CB 3.48 3.17 3.78 3.73
Si-O(Å) 1.62 1.63 1.60 1.60-1.62
B-O(Å) 1.49 1.46 1.47 1.37-1.47
Na-O(Å) 2.55 2.44 2.52 2.29-2.62
Si-Si(Å) 3.18 3.15 3.06 3.08
Si-O-Si 157.0 147.2 145.2 144-147
O-B-O 113.0 116.8 110.7 111.6-118.6
B-O-B 152.1 148.0 136.4 129.4-143
The pressure change with deposited energy is first presented, followed by results for all the
glasses before (BDC) and after (ADC) the DC series. This section is divided into subsections
dedicated to density, short-range order, intermediate-range order, network connectivity,
randomness and chemical mixing.
3.1.1. Evolution of pressure with deposited energy
Pressure change inside the glass with deposited energy (i.e. the energy deposited by the
cascade projectiles) was followed, see Figure 3. Evolution of pressure inside different SBN
(left) and SBNA (right) glass compositions with accumulation of cascades
. This shape of the curve − an initial rapid rise, followed by an increasingly gentler slope until
saturation is reached, for a deposited energy in the range 4-6 eV/atom − is qualitatively
consistent with experimental results [2] and with simulations performed by Kilymis et al.
[30,50,51]. For SBN-14 subjected to series of 120 (4 keV each) and 190 (0.8 keV each)
displacement cascades, saturation was attained at 6 eV/atom[50] and 8 eV/atom [10]
respectively, which corresponds to an equivalent dose of 4.3×1018
α/g (or 9.8×1021
keV/cm3).
Even though the total pressure accumulated at the end of the DC is different for different
compositions, the saturation threshold has been reached in all cases. In several studies on
doped, externally irradiated or simulated glasses [11,18,29], saturation in density, hardness
and other structural properties were observed around 2×1018
α/g or a nuclear dose of 1021
keV/cm3. This also implies that irrespective of the composition, system size and projectile
energy, the internal pressure threshold is reached in all cases around the similar accumulated
dose.
Figure 3. Evolution of pressure inside different SBN (left) and SBNA (right) glass
compositions with accumulation of cascades
3.1.2. Density
Densities before and after DC are plotted for SBN and SBNA glasses in Fig. 4 against K and
Kˊ ratio, respectively. In general, glass density increased with an increasing K or Kˊ up to a
value of approximately 3 and then decreased. SBN-955 (K=18 not shown in Fig. 4) had a
higher final density (2.77 g/cm3) than expected and there could be two possible explanations
for this: (a) the dependence of density on the quenching temperature, in particular some
fivefold-coordinated Si species were found, which might contribute to the high density (b) the
low number of Na and B atoms, which reduces the chances for [BO4]- + Na entities to form
and favoring BO3, in fact almost half of Na was used up in placing NBOs’ on Si and [3]
B.
Fig. 4 shows that density of each glass decreased after DC. Swelling of approximately 4-9%
was observed in SBN glasses and 4-5% in SBNA glasses. Similar swelling in borosilicate
glasses with the amplitude depending upon composition and irradiation conditions, has been
observed in studies with doped glasses (0.5-0.6% [11,13]), external irradiation (1.3-2.4%
[52,53]), and simulations (3-7% [28,30,54]).
Figure 4. Variation of density with K and K’ ratios for SBN (Green marker) and SBNA (Blue
marker), before (filled circles) and after (empty circles) series of 120 DC
3.1.3. Short and medium range orders
The effect of irradiation on short-range order can be evaluated by following the change in
coordination of glass formers. Each cation has a well-defined coordination state
corresponding to a tetrahedral or triangular local environment. These local entities act as
building blocks and exhibit order at the level of several associated ions, and this order extends
over around 1 nm defining the medium-range order. Any change in short- and medium-range
orders will also lead to change in radial distribution functions.
3.1.3.1. Coordination of glass formers
Si and Al essentially remained in tetrahedral coordination after DC (data not shown) as
evidenced by a change of less than 0.1% in the average Si and Al coordination. This result is
consistent with their high field strengths.
The non-linear trend observed in Fig. 5, for mean boron coordination (CB) with composition,
is in agreement with the Dell and Bray model and experimental results [55,56] . At high K
(K’) (low alkali concentration), Na primarily acts as charge compensator converting [3]
B to [4]
B, resulting in a gradual increase of CB. As K (K’) starts to decrease (increasing alkali
concentration), Na changes its role to a network modifier leading to low CB values.
After DC, CB decreased on an average by 2-4% for SBN and 3-4% for SBNA (Figure ). In
other words, a decrease in [4]
B of the order 12-22 % and 14-30% for SBN and SBNA,
respectively, was estimated. Studies conducted with external irradiation have shown a similar
decrease, 13-30%[52,53] in [4]
B, whereas, doped glasses [53] and glasses prepared with high
quenching rate [57,58] showed lesser decrease about 7% in [4]
B.
The decrease in CB has often been associated with the thermal history of glass preparation
[2,28,57]. In terms of ballistic effects, it can be understood as when the heavy projectile
traverses through the glass, it losses its energy by elastic collisions, which leads to a sharp rise
in temperature for a short time at the center of the cascade and eventually, the glass is rapidly
quenched under the influence of the surrounding material still at room temperature. It has
been observed that the chemical equilibrium described by Equation (7) is shifted towards the
right when temperature increases [58,59]. So [3]
B formation is favored in cascade core, and a
part of it , is retained due to the very high quench rate following the displacement cascade
explaining the [3]
B concentration increase under irradiation.
(7)
The decrease in CB also implies change in Na role from charge compensator to network
modifier. This might have an effect on chemical durability.
Figure 5. Variation of average B coordination with K and K’ ratio, for SBN (left) and SBNA
(right) respectively, before (solid line) and after (dotted line) series of DC.
3.1.3.2. Radial distribution functions (RDF)
RDF of Si-O and Al-O have single peaks with no significant change in their position but
slight broadening after series of DC, see Figs. 6 and 7, which is in agreement with no
significant coordination change observed for Si and Al and increase in disorder.
For the B-O RDF, the curve has two peaks depending on boron coordination, peak on left side
depicts [3]
B and other peak after that depicts [4]
B. Comparing SBN glasses in Fig. 6, top 2
plots, it can be seen for SBN-14 with Jolley’s potential, even before DC, majority of B are in [3]
B as compared to Kieu’s potential and the Si-O RDF has shifted slightly to the right. Other
four figures compare glasses with decreasing B2O3 or increasing SiO2 content (From SBN-433
to SBN-955).
After irradiation, in general, RDF for B-O can be seen shifting towards left. Decrease in
population of [4]
B and corresponding increase in [3]
B, leading into decrease in average bond
length, ,since
. [34]
Figure 6. Radial distribution functions for SBN glasses before (solid line) and after
(dotted line) series of DC. Top two are SBN-14 glass prepared with different
potentials and then glasses have been put in order of increasing silica content
All SBNA compositions possess 15%, B2O3 and although concentration of Na2O is increased
with increasing Al2O3 concentration such that all of Al is compensated first. Still, with
increasing Al2O3 or decreasing SiO2,
[4]B is seen decreasing ( see Fig.7). It is interesting to
note that [3]
B peak is prominent for SBNA-12, although, there is not much difference between
SBNA-12 and SBNA-10 composition. This may imply boron anomaly threshold for these
SBNA glasses and it will be seen in next section that there is indeed higher fraction of NBOs’
in this composition.
Figure 7. Radial distribution functions for SBNA glasses before (solid line) and
after (dotted line) series of DC. Except for SBNA-CJ2 (bottom) all other
compositions contain 15% B2O3 and have been put in order of increasing alumina
content
3.1.4. Network connectivity
Connectivity of a glassy network can be determined by the number and arrangement of
bridging and non-bridging bonds which link each of the building blocks to their neighbors.
Thus, entities Qn and Tn are analyzed to determine the connectivity of the network.
3.1.4.1. Qn and Tn Distribution
Figs. 8-11 can be correlated with section 3.2.3.1, wherein CB was seen to increase with K and
at low K ratio, Na was referred to as network modifier. Indeed at low K (or high alkali
concentration), there is higher fraction of low order Qn and Tn for both the type of glasses.
Comparing evolution after series of DC for each former. For Si (see Figs. 8 and 9), a decrease
in the range 3-7% (for SBN) and 1-4% (for SBNA) in Q4 and corresponding increase in Q3,
Q2 and Q1, indicating decrease in Si connectivity after series of DC. In a study by Peuget et.
al., NMR characterization of borosilicate glass irradiated by gold ions revealed slightly higher
proportion of Q3 after irradiation[52].
Figure 8. Variation of normalized Qn for Si
with K ratio for different SBN glasses, before
(solid line) and after (dotted line) series of
DC
Figure 5. Variation of normalized Qn for Si
with K’ ratio for different SBNA glasses,
before (solid line) and after (dotted line)
series of DC
For B (see Figs. 10 and 11), decrease in T3 was estimated, almost 2-6 % of T3 transforms to
T2 and T1, which means, a fraction of NBO is also placed on 3B entities. Looking at the Qn for
B, a decrease in Q4 by 1-4% and corresponding increase in Q3 is seen.
Figure 60.Variation of Qn and Tn for B with K ratio for different SBN glasses, before (solid
line) and after (dotted line) series of DC
Figure 71.Variation of Tn and Qn for B with K’ ratio for different SBNA glasses, before (solid
line) and after (dotted line) series of DC
Majority of Al were present as Q4 (not shown here), however there was a decrease in the
range 1.6-4.6% (depending upon composition) in Q4 and corresponding increase in Q3 and Q2
species after DC.
This increase in fraction of low order Qn and Tn entities can be correlated with the creation of
NBO, when BO4 transforms to BO3 under irradiation, see Equation 5. Available sodium,
which was earlier acting as a charge compensator near a [BO4]- unit, now changes its role and
acts as network modifier around the Si, B and Al formers. NBO distribution will be discussed
in next section.
3.1.4.2. NBO distribution
Before DC maximum number of NBOs’ on Si species (except for SBNA-12) followed by B
and then Al, see Figs. 12 and 13, and further after DC, there is an increase in NBOs’ on each
species (dashed lines). For Si, B and Al the percentage increase was in the range 13-40%, 17-
50% and 39-42%, respectively. A closer look at NBOs’ placed on [3]
B and [4]
B, Figs. 13 (b)
and 14 (b), show higher proportion of NBOs’ on [3]
B, especially at low K. This is in
agreement with Dell and Bray model for borosilicate glasses, as well as with other
experimental studies on borosilicate glasses [56]. Further, after series of DC slightly higher
fraction of NBOs’ was estimated on [3]
B entities.
Figure 8. Variation in number of, (a). NBO per Si or B species (b). NBO per trigonal
(3B) and tetrahedral (4B) species , with K ratio for SBN glasses, before (solid line) and
after (dotted line) series of DC
Figure 9. Variation in number of, (a). NBO per Si, B or Al species (b). NBO per trigonal
(3B) and tetrahedral (4B) species, with K’ ratio for SBNA glasses, before (solid line)
and after (dotted line) series of DC
3.1.5. Ring distribution
Though glasses are amorphous and expected to exhibit a high degree of disorder, within
themselves, it has been observed that there exist rings and chains of building blocks connected
in a manner closely found in crystals, but yet not extending over long range distances.
Distribution of such entities is said to provide a degree of intermediate range order to the
structure[2,60].
(a)
(b)
(a) (b)
The glass topologies have been studied by analyzing rings statistics using software R.I.N.G.S
[61]. The rings are composed of Si, Al and B connected by O atoms. For SBN glasses
(Fig.14), pristine or irradiated, it can be seen with increasing silica content there is narrowing
of distribution around 6,7,8,9 membered rings. Similarly for SBNA glasses (Fig.15), with
increasing alumina and decreasing silica, the distribution can be seen widening. This
underlines the strong order imposed by the silica in the glass network because of the very
rigid SiO4 entities.
The ring distributions for all the glasses broaden after DC (Figs.14 and 15). In general there is
an increase in number of small and large rings and decrease in number of intermediate rings.
Further, there is also decrease in total number of ring entities (by 4-6%), which is in
agreement with the depolymerisation induced by the NBO formation. It has been observed in
that after irradiation, there is broadening of ring distribution and an increase in population of 3
members’ ring [2,27,62]. In fact, this increase of structural disorder in the glass topologies can
be explained by the increase in fictive temperature induced by ballistic effects. This higher
entropy is again associated to the increase of the quench rate inside the cascade cores leading
to a less stable (i.e. characterized by a higher internal energy) final structure.
Figure 10. Ring size distribution for different SBN glasses before
(solid line) and after (dotted line) series of DC. From top, glasses
have been arranged in order of increasing silica content until
SBN-955, after that SBN-14 prepared with two different
potentials.
Figure 11. Ring size distribution for different SBNA glasses before
(solid line) and after (dotted line) series of DC. From top, glasses
have been arranged in order of increasing alumina content.
3.1.6. Randomness and chemical mixing
Structural modification in the glass network is also reflected by the angular distribution
function of triplets F-O-F’ (F and F’ are network formers Si, B, or Al). Changes in the width
of distribution can also give information on the degree of topological disorder in the network.
3.1.6.1. Angular distribution
With increasing silica content and respective decrease in boron content (See Fig.16), there is
sharpening of Si-O-Si peak, broadening of Si-O-B and diminishing B-O-B, as we go from
SBN-433 to SBN-955. After series of DC, broadening (confirmed by calculating FWHM, see
supporting information) and slight shift to the left of distribution function can be seen,
implying that there is an increase in topological disorder and decrease in mean angle. Average
decrease in Si-O-Si, Si-O-B and B-O-B was 2.41°, 1.977° and 3.28° respectively and average
increase in O-B-O angle by 0.78° was observed. This increase is due to the formation of three
coordinated B atoms.
For SBNA glasses (Fig.17) from SBNA-2 to SBNA-12, sharpening of Si-O-Al triplet angular
distribution and diminishing intensities for Si-O-Si and Si-O-B which is in agreement with
increasing alumina content. Further, after series of DC, Si-O-Si and Si-O-B can be clearly
seen broadening and shifting towards left, this change is not as prominent for Si-O-Al. The
mean decrease in Si-O-Si, Si-O-B, B-O-B, and Si-O-Al was 2.45°, 1.77°, 1.98° and 1.81°,
respectively and mean increase in O-B-O and Al-O-Al was 1.023° and 2.30°, respectively.
Decrease in mean Si-O-Si angle has been observed in- Cm doped ISG1 glass [52], gold
irradiated CJ1 glass [62] and other molecular dynamics simulations on borosilicate glasses
[11,28,29]. This decrease has been associated with the increasing depolymerization and
disorder (observed in previous sections), which leads to decrease in number of ring entities
and broadening of ring distribution. Similar effects on Si-O-Si angle have been observed in
glasses prepared at high quenching rate.
1 International Simple Glass [16.0B2O3, 12.6Na2O, 3.8Al2O3, 5.7CaO, 1.7ZrO2]
Figure 126. Angular distribution function for triplets for different SBN
glasses before (solid line) and after (dotted line) series of DC
Figure 13. Angular distribution function for triplets for different SBNA
glasses before (solid line) and after (dotted line) series of DC
3.1.6.2. Distribution of triplets
Following the method presented in section 2.4.2, number of triplets associated with two
coordinated oxygen (2O), assuming random mixing, were calculated. Comparing the shaded
region between solid lines (BDC) and dotted lines (ADC) (see Fig. 18), implying an increase
in random mixing even though the number of mixed triplets seems to decrease.
Figure 14. Comparison between number of triplets from random mixing and simulation.
Distribution of triplets before (solid lines) and after (dotted lines) in SBN (left) and SBNA
(Right) glasses with K and K’ ratio respectively. The shaded area between the solid lines is
higher than area between dotted lines
4. Discussion
3.1. Validity of potentials
All the glasses studied in the current paper were prepared with the potentials proposed by
Kieu and Deng for borosilicate and alumino-borosilicates, respectively. Considering the
selected preparation scheme, these potentials slightly overestimate the density but the other
investigated structural parameters were found to be in good agreement with experimental
results. Density and coordination of SBN-955 was not correctly simulated. It should be noted
that this composition lies outside the scope of Kieu’s potentials[33] although other reasons
such as high quenching rate and low probability of boron finding sodium for charge
compensation were also cited as possible reasons for these shifts. SBN-955 was prepared with
lower quenching rate (2.5K/ps). The estimation on density and boron coordination did not
show any significant improvement, implying inability to simulate high silica content glasses
may be an inherent drawback of this potential.
3.2. Ballistic effects and role of Al2O3
Even if in the scope of this study, SBNA glasses were found to be less impacted by ballistic
effects compared to SBN glasses, it is interesting to notice that both in borosilicate and
alumino silicate glasses, accumulation of displacement cascades lead to a new glass
configuration characterized by a lower density and a larger structural disorder, i.e. increase of
the width of the angular distributions and ring size distributions, increase of the chemical
mixing, increase in the NBO and 3O concentrations. These results are in agreement with
experimental ones and with previous molecular dynamics studies [11,27,29,50,53,54].
The Al2O3 addition does not modify qualitatively the ballistic effects but their intensity is
reduced. Table 3 compares various structural parameters between SBN and SBNA glasses
after series of DC. Average has been taken over all SBN or SBNA glasses for each parameter.
In glasses containing alumina, swelling, depolymerization and structural disordering are
lower. In fact more NBOs’ and 3O entities are formed after irradiation in SBN glasses.
Fig. 19, decrease in depolymerization after irradiation when the Al2O3/SiO2 ratio increases is
confirmed as less NBOs are formed around the Si and B local entities. Nevertheless the
number of NBOs introduced per Al does not change significantly.
Table 3.Comparison of structural parameters between SBN and SBNA glasses after series of
DC, the middle column between SBN and SBNA shows relation and last column presents the
trend followed after DC. Average has been taken over all glass compositions for SBN and
SBNA glasses
Parameter Average change with respect to initial value
SBN SBNA Trend
<Density> -0.148 > -0.128 Decrease
<CB> -0.112 < -0.143 Decrease
<SiOSi> -2.41 ~ -2.45 Decrease
<OSiO> -0.011 > -0.003 Decrease
<BOB> -3.287 > -1.98 Decrease
<OBO> +0.78 < +1.02 Increase
<SiOB> -1.978 > -1.775 Decrease
<O3BO> +0.079 ~ +0.069 Increase
<NBO> +661.39 > +639.59 Increase
<2O> -802.18 > -674.30 Decrease
<3O> +149.25 >> +34.90 Increase
Figure 15. Fractional increase after DC, in number of NBOs’
per species versus alumina to silica ratio.
3.3. How displacement cascade results will be used in Monte Carlo calculations of the
glass alteration
It has been marked by various studies that decrease in boron coordination and
depolymerisation in glasses, may affect the reactivity of the glass towards dissolution
[2,22,63–67]. Impact of other parameters such as evolution of angular and ring distribution,
on dissolution of glass, are still not well understood. On the other hand, it has been widely
accepted that leaching will be the most probable process of release of radionuclides from the
glass, and recent experiments with irradiated glasses show an increase in leaching rate and
alteration depth. In the ongoing study, effect of irradiation on glass leaching will be
investigated by considering in Monte Carlo simulations, boron coordination change,
depolymerisation, and angular and ring distribution change. This will be done first by
modifying the Monte Carlo network by changing the average B coordination or the
polymerization level, and second by playing on the different probabilities (probability for
dissolution and condensation) by replacing constant values by distributions of values.
5. Conclusions
To study ballistic effects in SiO2-B2O3-Na2O and SiO2-B2O3-Na2O-Al2O3 glasses, series of
displacement cascades were simulated for a large set of compositions. Qualitatively, the
borosilicate and alumino-borosilicate glasses behaved in the same way under the
accumulation of displacement cascades. A swelling is systematically observed and the
irradiated structures are characterized by a larger structural disorder and an increase of the
chemical mixing that approaches the random chemical mixing.
Nevertheless, it has been observed that the radiation effects are attenuated after the Al2O3
addition. The swelling and the structural changes are lower. This can be understood as the
consequence of the large strength of the AlO4 entities. During the quench after the radiation
effects, these entities rebuild very rapidly and in consequence the structural changes are
limited.
ACKNOWLEDGMENTS
We are grateful for the computing resources
References
[1] S. Gin, A. Abdelouas, L.J. Criscenti, W.L. Ebert, K. Ferrand, T. Geisler, M.T. Harrison, Y. Inagaki, S.
Mitsui, K.T. Mueller, J.C. Marra, C.G. Pantano, E.M. Pierce, J.V. Ryan, J.M. Schofield, C.I. Steefel,
J.D. Vienna, An international initiative on long-term behavior of high-level nuclear waste glass, Mater.
Today. 16 (2013) 243–248. doi:10.1016/j.mattod.2013.06.008.
[2] W.J. Weber, R.C. Ewing, C.A. Angell, G.W. Arnold, A.N. Cormack, J.M. Delaye, D.L. Griscom, L.W.
Hobbs, A. Navrotsky, D.L. Price, A.M. Stoneham, M.C. Weinberg, Radiation effects in glasses used for
immobilization of high-level waste and plutonium disposition, J. Mater. Res. 12 (1997) 1946–1978.
doi:10.1557/JMR.1997.0266.
[3] M. Tribet, S. Rolland, S. Peuget, V. Broudic, M. Magnin, T. Wiss, C. Jégou, Irradiation Impact on the
Leaching Behavior of HLW Glasses, Procedia Mater. Sci. 7 (2014) 209–215.
doi:10.1016/j.mspro.2014.10.027.
[4] S. Gin, Open Scientific Questions about Nuclear Glass Corrosion, Procedia Mater. Sci. 7 (2014) 163–
171. doi:10.1016/j.mspro.2014.10.022.
[5] J.D. Vienna, J. V. Ryan, S. Gin, Y. Inagaki, V.J. D., R.J. V., G. Stéphane, I. Yaohiro, J.D. Vienna, J. V.
Ryan, S. Gin, Y. Inagaki, Current Understanding and Remaining Challenges in Modeling Long‐Term
Degradation of Borosilicate Nuclear Waste Glasses, Int. J. Appl. Glas. Sci. 4 (2013) 283–294.
doi:10.1111/ijag.12050.
[6] C. Poinssot, S. Gin, Long-term Behavior Science: The cornerstone approach for reliably assessing the
long-term performance of nuclear waste, J. Nucl. Mater. 420 (2012) 182–192.
[7] S. Gin, P. Jollivet, M. Fournier, F. Angeli, P. Frugier, T. Charpentier, Origin and consequences of
silicate glass passivation by surface layers, Nat. Commun. 6 (2015) 6360.
[8] S. Gin, M. Collin, P. Jollivet, M. Fournier, Y. Minet, L. Dupuy, T. Mahadevan, S. Kerisit, J. Du,
Dynamics of self-reorganization explains passivation of silicate glasses, Nat. Commun. 9 (2018) 2169.
doi:10.1038/s41467-018-04511-2.
[9] J.-M. Delaye, D. Ghaleb, Dynamic processes during displacement cascades in oxide glasses: A
molecular-dynamics study, Phys. Rev. B. 61 (2000) 14481–14494. doi:10.1103/PhysRevB.61.14481.
[10] D.A. Kilymis, J.-M. Delaye, S. Ispas, Density effects on the structure of irradiated sodium borosilicate
glass: A molecular dynamics study, J. Non-Cryst. Solids. 432 (2016) 354.
doi:10.1016/J.JNONCRYSOL.2015.10.031.
[11] S. Peuget, J.-M. Delaye, C. Jégou, Specific outcomes of the research on the radiation stability of the
French nuclear glass towards alpha decay accumulation, J. Nucl. Mater. 444 (2014) 76–91.
doi:10.1016/J.JNUCMAT.2013.09.039.
[12] S. Gin, P. Jollivet, M. Tribet, S. Peuget, S. Schuller, Radionuclides containment in nuclear glasses: An
overview, Radiochim. Acta. 105 (2017) 927–959. doi:10.1515/ract-2016-2658.
[13] W.J. Weber, Radiation and Thermal Ageing of Nuclear Waste Glass, Procedia Mater. Sci. 7 (2014) 237–
246. doi:10.1016/J.MSPRO.2014.10.031.
[14] W.J. Weber, F.P. Roberts, A Review of Radiation Effects in Solid Nuclear Waste Forms, Nucl. Technol.
60 (1983) 178–198. doi:10.13182/NT83-A33073.
[15] S. Peuget, P.-Y. Noël, J.-L. Loubet, S. Pavan, P. Nivet, A. Chenet, Effects of deposited nuclear and
electronic energy on the hardness of R7T7-type containment glass, Nucl. Instruments Methods Phys.
Res. Sect. B Beam Interact. with Mater. Atoms. 246 (2006) 379–386.
[16] T. Advocat, P. Jollivet, J.. Crovisier, M. del Nero, Long-term alteration mechanisms in water for SON68
radioactive borosilicate glass, J. Nucl. Mater. 298 (2001) 55–62. doi:10.1016/S0022-3115(01)00621-3.
[17] S. Rolland, T. Magaly, J. Christophe, B. Véronique, M. Magali, P. Sylvain, W. Thierry, J. Arne, B.
Antoine, T. Pierre, 99Tc‐ and 239Pu‐Doped Glass Leaching Experiments: Residual Alteration Rate and
Radionuclide Behavior, Int. J. Appl. Glas. Sci. 4 (2013) 295–306. doi:10.1111/ijag.12051.
[18] S. Peuget, V. Broudic, C. Jégou, P. Frugier, D. Roudil, X. Deschanels, H. Rabiller, P.Y. Noel, Effect of
alpha radiation on the leaching behaviour of nuclear glass, J. Nucl. Mater. 362 (2007) 474–479.
[19] S. Mougnaud, M. Tribet, J.-P. Renault, P. Jollivet, G. Panczer, T. Charpentier, C. Jégou, Effect of low
dose electron beam irradiation on the alteration layer formed during nuclear glass leaching, J. Nucl.
Mater. 482 (2016) 53–62. doi:10.1016/J.JNUCMAT.2016.10.008.
[20] D.M. Wellman, J.P. Icenhower, W.J. Weber, Elemental dissolution study of Pu-bearing borosilicate
glasses, J. Nucl. Mater. 340 (2005) 149–162. doi:10.1016/j.jnucmat.2004.10.166.
[21] X. Deschanels, S. Peuget, J.N. Cachia, T. Charpentier, Plutonium solubility and self-irradiation effects in
borosilicate glass, Prog. Nucl. Energy. 49 (2007) 623–634. doi:10.1016/J.PNUCENE.2007.05.001.
[22] S. Mougnaud, M. Tribet, J.-P. Renault, S. Gin, S. Peuget, R. Podor, C. Jégou, Heavy ion radiation ageing
impact on long-term glass alteration behavior, J. Nucl. Mater. 510 (2018) 168–177.
doi:10.1016/J.JNUCMAT.2018.07.046.
[23] S. Kerisit, E.M. Pierce, Monte Carlo simulations of the dissolution of borosilicate glasses in near-
equilibrium conditions, J. Non. Cryst. Solids. 358 (2012) 1324–1332.
doi:10.1016/j.jnoncrysol.2012.03.003.
[24] F. Devreux, A. Ledieu, P. Barboux, Y. Minet, Leaching of borosilicate glasses. II. Model and Monte-
Carlo simulations, J. Non. Cryst. Solids. 343 (2004) 13–25.
[25] S. Kerisit, E.M. Pierce, J. V. Ryan, Monte Carlo simulations of coupled diffusion and surface reactions
during the aqueous corrosion of borosilicate glasses, J. Non. Cryst. Solids. 408 (2015) 142–149.
[26] J.-M. Delaye, D. Ghaleb, Combining two types of molecular dynamics for rapid computation of high-
energy displacement cascades. II. Application of the method to a 70-keV cascade in a simplified nuclear
glass, Phys. Rev. B. 71 (2005) 224204. doi:10.1103/PhysRevB.71.224204.
[27] D. Kilymis, A. Faivre, T. Michel, S. Peuget, J.-M. Delaye, J. Delrieu, M. Ramonda, S. Ispas, Raman
spectra of indented pristine and irradiated sodium borosilicate glasses, J. Non. Cryst. Solids. 464 (2017)
5–13. doi:10.1016/j.jnoncrysol.2017.03.012.
[28] J.-M. Delaye, S. Peuget, G. Calas, L. Galoisy, Comparative effects of thermal quenching and ballistic
collisions in SiO2–B2O3–Na2O glass, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact.
with Mater. Atoms. 326 (2014) 256–259. doi:10.1016/J.NIMB.2013.10.061.
[29] J.-M. Delaye, S. Peuget, G. Bureau, G. Calas, Molecular dynamics simulation of radiation damage in
glasses, J Non-Cryst. Solids. 357 (2011) 2763.
[30] J.M. Delaye, D. Ghaleb, Volume change origin in glasses subjected to ballistic collisions: Molecular
dynamics simulations, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms.
191 (2002) 10–16. doi:10.1016/S0168-583X(02)00505-0.
[31] E.A. Maugeri, S. Peuget, D. Staicu, A. Zappia, C. Jegou, T. Wiss, Calorimetric Study of Glass Structure
Modification Induced by α Decay, J. Am. Ceram. Soc. 95 (2012) 2869–2875. doi:10.1111/j.1551-
2916.2012.05304.x.
[32] W.J. Dell, P.J. Bray, S.Z. Xiao, 11B NMR studies and structural modeling of Na2O-B2O3-SiO2 glasses
of high soda content, J. Non. Cryst. Solids. 58 (1983) 1–16. doi:10.1016/0022-3093(83)90097-2.
[33] L. Deng, J. Du, Development of effective empirical potentials for molecular dynamics simulations of the
structures and properties of boroaluminosilicate glasses, J. Non. Cryst. Solids. 453 (2016) 177–194.
doi:10.1016/J.JNONCRYSOL.2016.09.021.
[34] L.-H. Kieu, J.-M.J.-M. Delaye, L. Cormier, C. Stolz, Development of empirical potentials for sodium
borosilicate glass systems, J. Non-Cryst. Solids. 357 (2011) 3313. doi:10.1016/j.jnoncrysol.2011.05.024.
[35] Y.H. Yun, P.J. Bray, Nuclear magnetic resonance studies of the glasses in the system Na2O-B2O3-SiO2,
J. Non-Cryst. Solids. 27 (1978) 363. doi:10.1016/0022-3093(78)90020-0.
[36] K. Jolley, R. Smith, K. Joseph, Borosilicate glass potentials for radiation damage simulations, Nucl.
Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 352 (2015) 140–144.
doi:10.1016/J.NIMB.2014.12.024.
[37] J.-M. Delaye, V. Louis-Achille, D. Ghaleb, Modeling oxide glasses with Born–Mayer–Huggins
potentials: Effect of composition on structural changes, J. Non. Cryst. Solids. 210 (1997) 232–242.
doi:10.1016/S0022-3093(96)00604-7.
[38] L. Cormier, D. Ghaleb, D.R. Neuville, J.-M. Delaye, G. Calas, Chemical dependence of network
topology of calcium aluminosilicate glasses: a computer simulation study, J. Non. Cryst. Solids. 332
(2003) 255–270. doi:10.1016/J.JNONCRYSOL.2003.09.012.
[39] L. Huang, J. Kieffer, Thermomechanical anomalies and polyamorphism in B2O3 glass: A molecular
dynamics simulation study, Phys. Rev. B. 74 (2006) 224107. doi:10.1103/PhysRevB.74.224107.
[40] L. Huang, J. Kieffer, Molecular dynamics study of cristobalite silica using a charge transfer three-body
potential: Phase transformation and structural disorder, J. Chem. Phys. 118 (2003) 1487–1498.
doi:10.1063/1.1529684.
[41] S. H Garofalini, H.M. Tai, U. Joelyn, Simulations of the surfaces of soda lime aluminoborosilicate
glasses exposed to water, J. Am. Ceram. Soc. 101 (2017) 1135–1148. doi:10.1111/jace.15237.
[42] H. Inoue, A. Masuno, Y. Watanabe, Modeling of the Structure of Sodium Borosilicate Glasses Using
Pair Potentials, J. Phys. Chem. B. 116 (2012) 12325–12331. doi:10.1021/jp3038126.
[43] A.F. Alharbi, K. Jolley, A.J. Archer, J.K. Christie, A new potential for radiation studies of borosilicate
glass, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 393 (2017) 73–
76. doi:10.1016/j.nimb.2016.12.007.
[44] P. Stoch, A. Stoch, Structure and properties of Cs containing borosilicate glasses studied by molecular
dynamics simulations, J. Non. Cryst. Solids. 411 (2015) 106–114. doi:10.1016/j.jnoncrysol.2014.12.029.
[45] I.T. Todorov, W. Smith, K. Trachenko, M.T. Dove, DL_POLY_3: new dimensions in molecular
dynamics simulations via massive parallelism, J. Mater. Chem. 16 (2006) 1911. doi:10.1039/b517931a.
[46] E. de Leeuw ,S;Perram ,J; Smith, Simulation of electrostatic systems in periodic boundary conditions. I.
Lattice sums and dielectric constants, Proc. R. Soc. London. A. Math. Phys. Sci. 373 (1980) 27 LP-56.
[47] J.F. Ziegler, M.D. Ziegler, J.P. Biersack, SRIM – The stopping and range of ions in matter (2010), Nucl.
Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 268 (2010) 1818–1823.
doi:10.1016/J.NIMB.2010.02.091.
[48] J.B. Ziegler, J.P. Biersack, U. Littmark, The Stopping Power and Ranges of Ions in Matter, 1985.
[49] J.-M. Delaye, L. Cormier, D. Ghaleb, G. Calas, Investigation of multicomponent silicate glasses by
coupling WAXS and molecular dynamics, J. Non. Cryst. Solids. 293–295 (2001) 290–296.
doi:10.1016/S0022-3093(01)00680-9.
[50] D.A. Kilymis, J.-M. Delaye, Nanoindentation studies of simplified nuclear glasses using molecular
dynamics, J. Non-Cryst. Solids. 401 (2014) 147.
[51] D.A. Kilymis, J.-M. Delaye, Deformation mechanisms during nanoindentation of sodium borosilicate
glasses of nuclear interest, J. Chem. Phys. 141 (2014) 14504.
[52] C. Mendoza, E.A. Maugeri, J.M. Delaye, R. Caraballo, T. Charpentier, M. Tribet, O. Bouty, C. Jégou,
Alpha Decays Impact on Nuclear Glass Structure, Procedia Mater. Sci. 7 (2014) 252–261.
doi:10.1016/j.mspro.2014.10.033.
[53] T. Charpentier, L. Martel, A.H. Mir, J. Somers, C. Jégou, S. Peuget, Self-healing capacity of nuclear
glass observed by NMR spectroscopy, Sci. Rep. 6 (2016) 25499.
[54] D.A. Kilymis, J.-M. Delaye, S. Ispas, Nanoindentation of the pristine and irradiated forms of a sodium
borosilicate glass: Insights from molecular dynamics simulations, J. Chem. Phys. 145 (2016) 44505.
doi:10.1063/1.4959118.
[55] M. Barlet, J.-M. Delaye, B. Boizot, D. Bonamy, R. Caraballo, S. Peuget, C.L. Rountree, From network
depolymerization to stress corrosion cracking in sodium-borosilicate glasses: Effect of the chemical
composition, J. Non. Cryst. Solids. 450 (2016) 174–184. doi:10.1016/j.jnoncrysol.2016.07.017.
[56] M. Barlet, A. Kerrache, J.-M. Delaye, C.L. Rountree, SiO2–Na2O–B2O3 density: A comparison of
experiments, simulations, and theory, J. Non. Cryst. Solids. 382 (2013) 32–44.
doi:10.1016/J.JNONCRYSOL.2013.09.022.
[57] F. Angeli, O. Villain, S. Schuller, T. Charpentier, D. de Ligny, L. Bressel, L. Wondraczek, Effect of
temperature and thermal history on borosilicate glass structure, Phys. Rev. B. 85 (2012) 54110.
doi:10.1103/PhysRevB.85.054110.
[58] J.F. Stebbins, S.E. Ellsworth, Temperature Effects on Structure and Dynamics in Borate and Borosilicate
Liquids: High-Resolution and High-Temperature NMR Results, J. Am. Ceram. Soc. 79 (1996) 2247–
2256. doi:10.1111/j.1151-2916.1996.tb08969.x.
[59] J. Wu, J.F. Stebbins, Quench rate and temperature effects on boron coordination in aluminoborosilicate
melts, J. Non. Cryst. Solids. 356 (2010) 2097–2108. doi:10.1016/J.JNONCRYSOL.2010.08.015.
[60] J.E. Shelby, Introduction to Glass Science and Technology, Royal Society of Chemistry, UK, 2005.
[61] S. Le Roux, P. Jund, Ring statistics analysis of topological networks: New approach and application to
amorphous GeS2 and SiO2 systems, Comput. Mater. Sci. 49 (2010) 70.
doi:10.1016/J.COMMATSCI.2010.04.023.
[62] S. Peuget, E.A. Maugeri, T. Charpentier, C. Mendoza, M. Moskura, T. Fares, O. Bouty, C. Jégou,
Comparison of radiation and quenching rate effects on the structure of a sodium borosilicate glass, J.
Non-Cryst. Solids. 378 (2013) 201. doi:10.1016/j.jnoncrysol.2013.07.019.
[63] M. Ren, L. Deng, J. Du, Bulk, surface structures and properties of sodium borosilicate and
boroaluminosilicate nuclear waste glasses from molecular dynamics simulations, J. Non. Cryst. Solids.
476 (2017) 87–94. doi:10.1016/j.jnoncrysol.2017.09.030.
[64] G.S. Frankel, J.D. Vienna, J. Lian, J.R. Scully, S. Gin, J. V Ryan, J. Wang, S.H. Kim, W. Windl, J. Du,
A comparative review of the aqueous corrosion of glasses, crystalline ceramics, and metals, Npj Mater.
Degrad. 2 (2018) 15. doi:10.1038/s41529-018-0037-2.
[65] F. Bouyer, G. Geneste, S. Ispas, W. Kob, P. Ganster, Water solubility in calcium aluminosilicate glasses
investigated by first principles techniques, J. Solid State Chem. 183 (2010) 2786–2796.
doi:10.1016/j.jssc.2010.08.031.
[66] G. Geneste, F. Bouyer, S. Gin, Hydrogen–sodium interdiffusion in borosilicate glasses investigated from
first principles, J. Non. Cryst. Solids. 352 (2006) 3147–3152. doi:10.1016/j.jnoncrysol.2006.04.023.
[67] B.C. Bunker, G.W. Arnold, D.E. Day, P.J. Bray, The effect of molecular structure on borosilicate glass
leaching, J. Non. Cryst. Solids. 87 (1986) 226–253. doi:10.1016/S0022-3093(86)80080-1.